OBJECTIVE: This analysis examines the structure and emergence of performance-based risk-sharing agreements from a theoretical economic perspective based on an analysis of the incentives facing manufacturers and payers--both public and private. BACKGROUND: Performance-based risk-sharing agreements are characterized as: 1) having a program of data collection to assess performance in terms of health outcomes; 2) linking price and/or revenue by formula to results; 3) being about performance in terms of health outcomes and cost-effectiveness rather than budget impact; and 4) providing a different distribution of risk between payer and manufacturer to that of a "conventional" transaction. METHODS: We review four related economics literatures that provide relevant insights to understanding these agreements: value of information, real option theory, money-back guarantees, and portfolio theory. We then integrate these literatures in a conceptual analysis of the incentives, constraints, and benefits and costs to manufacturers, payers, and patients to establish these agreements. RESULTS: Our overall finding is that cost and practicality of evidence collection are critical factors in deciding whether to set up a performance-based risk-sharing agreement. There are four important points in regard to feasibility: 1) Transaction costs are a key factor; 2) Evidence collection costs are the main part of transaction costs; and 3) It only makes sense to incur these costs if the uncertainties that can be "resolved" by the evidence are important, and 4) If evidence collection is to occur, then it makes sense to do this in a pre-agreed risk-sharing framework, i.e., this is preferable to a general understanding that there will be a renegotiation once the evidence has been collected. CONCLUSION: Finally, we ask: is performance-based payment for drugs or other medical products, a concept whose time has come or will it be another flash in the pan? The answer to this is not yet clear.